Hi all,
Can someone help me translate this TS stochastic code to AB? I went to the AFL
library and found various stochastics indicators but none match the TS
indicator. When I did the comparison, my data was the same for both TS
and AB. It is just that the AB indicator algorithm is different.
The TS function is in BLUE while the TS indicator is in RED.
thanks
{ Stochastic multiple-output
"series" function with choice of Original or Legacy
smoothing; see MULTIPLE-OUTPUT FUNCTIONS note below }
inputs:
PriceH( numericseries ),
PriceL( numericseries ),
PriceC( numericseries ),
StochLength( numericsimple ),
Length1( numericsimple ), { this input assumed to be a
constant > 0; used to slow
FastK to FastD = SlowK }
Length2( numericsimple ), { this input assumed to be a
constant > 0; used to slow
FastD to SlowD }
SmoothingType( numericsimple ), { this input is assumed to
be constant; pass in 1
for Original, 2 for Legacy }
oFastK( numericref ),
oFastD( numericref ),
oSlowK( numericref ),
oSlowD( numericref ) ;
variables:
LL( 0 ), { lowest low }
HH( 0 ), { highest high }
Num1( 0 ), { numerator 1 }
Den1( 0 ), { denominator 1 }
Num2( 0 ), { numerator 2 }
Den2( 0 ), { denominator 2 }
BarsToGo1( 0 ),
BarsToGo2( 0 ) ;
Stochastic = 1 ;
LL = Lowest( PriceL, StochLength ) ;
HH = Highest( PriceH, StochLength ) ;
Num1 = PriceC - LL ;
Den1 = HH - LL ;
if Den1 > 0 then
oFastK = Num1 / Den1 * 100
else
begin
oFastK = 0 ;
Stochastic = -1 ;
end ;
if SmoothingType = 1 then { Original }
begin
BarsToGo1 = Length1 - CurrentBar ;
if BarsToGo1 > 0 and CurrentBar > 0 then
begin
{ use approximate
"backpropagated" calculations until we have enough data }
Num2 = ( Cum( Num1 ) + BarsToGo1 * Num1[
CurrentBar - 1 ] ) / Length1 ;
Den2 = ( Cum( Den1 ) + BarsToGo1 * Den1[
CurrentBar - 1 ] ) / Length1 ;
end
else
begin
Num2 = Average( Num1, Length1 ) ;
Den2 = Average( Den1, Length1 ) ;
end ;
if Den2 > 0 then
oFastD = Num2 / Den2 * 100
else
begin
oFastD = 0 ;
Stochastic = -1 ;
end ;
BarsToGo2 = Length2 - CurrentBar ;
if BarsToGo2 > 0 and CurrentBar > 0 then
{ use approximate
"backpropagated" calculations until we have enough data }
oSlowD = ( Cum( oFastD ) + BarsToGo2 *
oFastD[ CurrentBar - 1 ] ) / Length2
else
oSlowD = Average( oFastD, Length2 ) ;
end
else if SmoothingType = 2 then { Legacy }
begin
oFastD = XAverage( oFastK, Length1 ) ;
oSlowD = XAverageOrig( oFastD, Length2 ) ;
end ;
oSlowK = oFastD ;
inputs:
AlertEndOfBar(true),
PriceH( High),
PriceL( Low),
PriceC( Close),
StochLength( 6),
SmoothingLength1( 3), { used to slow FastK to FastD = SlowK }
SmoothingLength2( 3), { used to slow FastD to SlowD }
SmoothingType( 1), { pass in 1 for Original, 2 for Legacy }
OverSoldC ( 20),
OverBoughtC ( 80),
MidLine ( 50),
OverSoldA ( 35),
OverBoughtA ( 65),
upColour(Blue),
downColour(Red),
colourDeltaBar(1);
variables:
oFastK( 0 ),
oFastD( 0 ),
oSlowK( 0 ),
oSlowD( 0 );
Value1 = Stochastic(
PriceH,
PriceL,
PriceC,
StochLength,
SmoothingLength1,
SmoothingLength2,
SmoothingType,
oFastK,
oFastD,
oSlowK,
oSlowD ) ;
Plot1( oSlowD, "SlowD") ;
Plot2 (OverSoldC, "OverSldC");
Plot3(OverBoughtC, "OverBotC");
Plot4 (OverSoldA, "OverSldA");
Plot5(OverBoughtA, "OverBotA");
Plot6(MidLine, "MidLine");
{ Color criteria }
if (oSlowD > oSlowD[1]) then
SetPlotColor[colourDeltaBar](1, upColour)
else if (oSlowD < oSlowD[1]) then
SetPlotColor[colourDeltaBar](1, downColour);
